4500
Organometallics 2009, 28, 4500–4506 DOI: 10.1021/om900363r
Niobium/Rhodium Bimetallic Complexes: Synthesis, Structure, and Catalytic Hydrosilylation of Acetophenone and Benzaldehyde Somying Leelasubcharoen,† Pavel A. Zhizhko,† Lyudmila G. Kuzmina,‡ Andrei V. Churakov,‡ Judith A. K. Howard,§ and Georgii I. Nikonov*,† †
Chemistry Department, Brock University, 500 Glenridge Avenue, St. Catharines, L2S 3A1, Ontario, Canada, ‡Institute of General and Inorganic Chemistry RAS, Leninskii Prosp. 31, 119991 Moscow, Russian Federation, and §Chemistry Department, University of Durham, South Road, Durham DH1 3LE, U.K. Received May 7, 2009
Reactions of a new imido phosphido complex, Cp2Nb(dNBut)(PPh2) (1), with olefin complexes of rhodium afford two types of compounds. With [(μ-Cl)Rh(C2H4)2]2, an imido/phosphido-bridged complex, Cp(Cl)Nb(μ-NBut)(μ-PPh2)RhCp (2), characterized by NMR and X-ray diffraction, was prepared. Formation of 2 represents a rare example of a Cp/Cl exchange reaction in organometallic chemistry. When 1 reacts with [(μ-Cl)Rh(COD)]2, the target phosphido-bridged bimetallic complex Cp2Nb(dNBut)(μ-PPh2)Rh(Cl)(COD) (4) is formed. The structure of this product was established by NMR and an X-ray diffraction study. Both 2 and 4 are catalyst precursors for hydrosilylation of benzaldehyde and acetophenone. The activity of the neutral complex 4 exceeds that of the cation [Cp2Nb(NBut)(PPh2)Rh(COD)]þ derived from 4 by chloride anion abstraction.
Introduction Heterobimetallic complexes1 have recently received renewed attention because the cooperative action of both metal atoms allows them to activate robust bonds and mediate complex transformations.2,3 All successful examples of active bimetallic systems are almost exclusively based on metals from the right (Rh, Pt, Cu, etc.) or middle part of transition series.3 Despite the fact that complexes featuring both early and late transition metals, i.e., early late heterobimetallics (ELHB), have been the subject of intensive research over the last 25 years,1 relatively little is known about their catalytic activity.1b,4 Previous work in this area was primarily based on the idea that electron-deficient early transition metals should serve as intramolecular Lewis acidic centers of ELHB, assisting the transformation of organic molecules such as carbonyls in the coordination sphere of late transition metals (LTM), thus facilitating catalysis.1a,4c For this reason, the earlier design of most ELHB complexes *To whom correspondence should be addressed. Tel: (905)6885550, ext 3350. Fax: (905)9328846. E-mail:
[email protected]. (1) Reviews: (a) Stephan, D. W. Coord. Chem. Rev. 1989, 95, 41–107. (b) Wheatley, N.; Kalck, P. Chem. Rev. 1999, 99, 3379–3419. (c) Gade, L. H. Angew. Chem., Int Ed. 2000, 39, 2659–2678. (2) (a) Rowsell, B. D.; McDonald, R.; Ferguson, M. J.; Cowie, M. Organometallics 2003, 22, 2944–2945. (b) Torkelson, J. R.; Antwi-Nsiah, F. H.; McDonald, R.; Cowie, M.; Pruis, J.; Jalkanen, K. J.; DeKock, R. L. J. Am. Chem. Soc. 1999, 121, 3666–3683. (c) Trepanier, S. J.; Sterenberg, B. T.; McDonald, R.; Cowie, M. J. Am. Chem. Soc. 1999, 121, 2613–2614. (3) Alvarez, C. M.; Alvarez, M. A.; Garcı´ a, M. E.; Garcı´ a-Viv o, D.; Ruiz, M. A. Organometallics 2005, 24, 4122–4124. (4) (a) Gelmini, L.; Stephan, D. W. Organometallics 1988, 7, 849. (b) Hostetler, M.; Butts, M.; Bergman, R. Organometallics 1993, 12, 65–75. (c) Hernandez-Gruel, M. A. F.; Perez-Torrente, J. J.; Ciriano, M. A.; Rivas, A. B.; Lahoz, F. J.; Dobrinovitch, I. T.; Oro, L. A. Organometallics 2003, 22, 1237–1249. (d) Cornelissen, C.; Erker, G.; Kehr, G.; Fr€ohlich, R. Organometallics 2005, 24, 214–225. pubs.acs.org/Organometallics
Published on Web 07/13/2009
was based on unsaturated group 4 metal moieties,1,4 and relatively little is known about the behavior of group 5 derivatives.4b,5 Another approach to a rational design of a bimetallic catalyst consists in the incorporation of one of the metal atoms into the ligand framework of the other catalytically active center, which is usually a LTM center. This can facilitate chemical transformation on the LTM by means of fine adjustment of ligand bonding capabilities. Such an approach has previously recommended itself in the catalytic Suzuki-Miyaura reaction mediated by phosphido-bridged Re/Pd and Ru/Pd complexes.6 In this case, the polarization of the Mδþ-Pδ- bond (M = Re or Ru) makes the phosphorus atom a better donor to the palladium atom. One can expect that metallophosphides, MPR2, based on the electropositive early transition metals will result in even greater polarization of the metal-phosphorus bond,6a which will further increase the bonding capabilities of the phosphido ligand toward a catalytically active late transition metal. Surprisingly, although a number of phosphido-bridged ELHBs are known,1a,1b,7 their application to organic synthesis and catalysis is scarce.4a (5) (a) Barre, C.; Boudot, P.; Kubicki, M.; Moı¨ se, C. Inorg. Chem. 1995, 34, 284. (b) Oudet, P.; Bonnet, G.; Moïse, C. Polyhedron 1995, 14, 2173. (c) Oudet, P.; Kubicki, M.; Moïse, C. Organometallics 1994, 13, 4278. (e) Boni, G.; Blacque, O.; Sauvageot, P.; Poujaud, N.; Moïse, C.; Kubicki, M. M. Polyhedron 2002, 21, 371. (6) (a) Eichenseher, S.; Kromm, K.; Delacroix, O.; Gladysz, J. A. Chem. Commun. 2002, 1046. (b) Planas, J. G.; Hampel, F.; Gladysz, J. A. Chem.;Eur. J. 2005, 11, 1402. (7) (a) Lindenberg, F.; Shribman, T.; Sieler, J.; Hey-Hawkins, E.; Eisen, M. S. J. Organomet. Chem. 1996, 515, 19. (b) Shribman, T.; Kurz, S.; Senff, U.; Lindberg, F.; Hey-Hawkins, E.; Eisen, M. S. J. Mol. Catal. A: Chem. 1998, 129, 191. (c) Wenzel, B.; L€onnecke, P.; Hey-Hawkins, E. Organometallics 2002, 21, 2070. r 2009 American Chemical Society
Article
Organometallics, Vol. 28, No. 15, 2009
Hemilabile ligands play a prominent role in modern catalysis.8 They usually feature a soft center (usually based on C, S, or P), which ensures strong binding to the metal, and a hard center (O or N based), which can bind loosely to a vacant metal site, thus creating a masked form of an unsaturated complex. We suggested that the attachment of hard and soft ligands such as imido and phosphido to an early transition metal would give rise to a new type of “spectator” metalloligand9 that could be used as a hemilabile ligand to a catalytically active late transition metal. Herein we report the application of this concept to the preparation of imido phosphido niobocene Cp2Nb(dNBut)(PPh2) (1) and the synthesis of some Nb/Rh heterobimetallics derived therefrom. In the course of our studies, we encountered an unusual exchange of Cp and chloride ligands between two metal centers. Application of the new Nb/Rh ELHB to catalytic hydrosilylation of acetophenone and benzaldehyde is also reported.
Results and Discussion Preparation of Cp2Nb(dNBut)(PPh2) (1). Our initial strategy to prepare complex 1 was based on the insertion/deprotonation method developed in this laboratory10 and by the group of Moı¨ se.11 Namely, we anticipated that the reaction between Cp2Nb(dNBut)(H)12 and ClPR2 would give the insertion product [Cp2Nb(dNBut)(PHR2)]Cl, which could be then deprotonated to furnish the phosphide 1. Related cationic P-H phosphine complexes have been previously prepared by reactions of Cp2MH3 and Cp2MH(L) (M = Nb, Ta; L = two-electron ligand) with chlorophosphines.10,11,13 In reality, it happens that Cp2Nb(dNBut)(H) reacts with ClPEt2 to give Cp2Nb(dNBut)(Cl) and HPEt2. We attribute the difference in chemical behavior between our imido hydride system and the previously studied trihydrides and monohydrides to the ease of phosphine displacement from the intermediate [Cp2Nb(dNBut)(PHR2)]þ, which generates an unsaturated intermediate, [Cp2Nb(dNBut)]þ, amenable to coupling with the chloride. Eventually, complex 1 was prepared in moderate yield by reacting Cp2Nb(dNBut)(Cl) with LiPPh2 3 Et2O in ether (eq 1). 1 was isolated as a yellow crystalline solid and characterized by multinuclear NMR spectroscopy. Similar (8) (a) Braunstein, P.; Naud, F. Angew. Chem., Int. Ed. 2001, 40, 680. (b) Braunstein, P. J. Organomet. Chem. 2004, 689, 3953. (c) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem. Rev. 2000, 100, 1169. (d) Helmchen, G.; Pfaltz, A. Acc. Chem. Res. 2000, 33, 336–345. (e) Lindner, E.; Pautz, S.; Haustein, M. Coord. Chem. Rev. 1996, 155, 145. (f) Bader, A.; Lindner, E. Coord. Chem. Rev. 1991, 108, 27. (g) Jimenez, M. V.; Perez-Torrente, J. J.; Bartolome, M. I.; Gierz, V.; Lahoz, F. J.; Oro, L. A. Organometallics 2008, 27, 224–234. (h) Grushin, V. V. Organometallics 2001, 20, 3950–3961. (i) Bassetti, M. Eur. J. Inorg. Chem. 2006, 4473. (9) Delacroix, O.; Gladysz, J. A. Chem. Commun. 2003, 665. (10) (a) Nikonov, G. I.; Lemenovskii, D. A.; Lorberth, J. Organometallics 1994, 13, 3127. (b) Nikonov, G. I.; Kuzmina, L. G.; Mountford, P.; Lemenovskii, D. A. Organometallics 1995, 14, 3588. (c) Nikonov, G. I.; Grishin, Y. K.; Lemenovskii, D. A.; Kazennova, N. B.; Kuzmina, L. G.; Howard, J. A. K. J. Organomet. Chem. 1997, 547, 183. (11) (a) Bonnet, G.; Lavastre, O.; Boni, G.; Leblanc, J. C.; Moı¨ se, C. C. R. Acad. Sci. Paris, Ser. 2 1994, 319, 1293. (b) Bonnet, G.; Lavastre, O.; Leblanc, J.-C.; Moïse, C. New J. Chem. 1988, 12, 551. (c) Barre, C.; Kubicki, M. M.; Leblanc, J.-C.; Moïse, C. Inorg. Chem. 1990, 29, 5244. (d) Bonnet, G.; Kubicki, M. M.; Moïse, C.; Lazzaroni, R.; Salvador, P.; Vitulli, G. Organometallics 1992, 11, 964. (12) Chernega, A. N.; Green, M. L. H.; Suarez, A. G. Dalton Trans. 1993, 3031. (13) Antinolo, A.; Carrillo-Hermosilla, F.; Fernandez-Baeza, J.; Garcia-Yuste, S.; Otero, A.; Sanchez-Prada, J.; Villasenor, E. Eur. J. Inorg. Chem. 2000, 1437.
4501
chloride substitution by MeLi to give Cp2Nb(dNBut)(Me) has been previously described.12 Although the 31P NMR signal of 1 was not detectable due to the fast relaxation on the 93 Nb nucleus (spin I = 9/2, natural abundance 100%), its position can be established from a cross-peak in the 1H-31P HSQC NMR between the ortho-protons of the phenyl group and the phosphorus signal at δ = -0.2 ppm.
Preparation of Bimetallic Complexes: Reaction with [(μCl)(Rh(C2H4)2]2. Diolefin complexes of Rh(I) in combination with phosphine, NHC carbenes, or related two-electron donors are common catalysts of hydrosilylation of carbonyls.14,15 Several families of related hemilabile ligands were applied in this catalytic system.15c,15d Bearing this in mind, we targeted the preparation of an adduct between the imido-phosphide 1 and [(μ-Cl)(Rh(C2H4)2]2. To our surprise, the reaction between 1 and 0.5 equiv of [(μ-Cl)(Rh(C2H4)2]2 in toluene resulted in the exchange compound Cp((Cl)Nb(μ-NBut)(μ-PPh2)RhCp (2) featuring the Cp ligand at rhodium and the chloride at niobium (eq 2). This is a very rare example of Cp ring substitution at an early transition metal center.16 The 1H NMR spectrum of 2 exhibits two Cp resonances, 4.88 and 5.42 ppm, of equal intensity that integrate as five protons each relative to the signals of the Ph and But groups. Correlation of these proton signals with the 13C NMR signals at 87.3 and 104.4 ppm, respectively, supports their assignment. The 1H signal of the But group at 1.72 ppm in 2 is shifted downfield relative to the corresponding resonance in 1 (1.02 ppm), indicating the presence of a bridging imido group in the former.
In order to get further insight into the structure of 2, a crystal structure determination was carried out (Figure 1). (14) (a) Ojima, I. In The Chemistry of Organic Silicon Compounds; Patai, S., Rappoport, Z., Eds.; Wiley: New York, 1989; Chapter 25. (b) Marciniec, B. Comprehensive Handbook on Hydrosilylation; Pergamon: Oxford, 1992. (c) Carpentier, J.-F.; Bette, V. Current Org. Chem. 2002, 6, 913. (d) Riant, O.; Mostefai, N.; Courmarcel, J. Synthesis 2004, 18, 2943. (15) (a) Ojima, I.; Kumagai, M.; Kogure, T.; Horiuchi, S.; Sato, T. J. Organomet. Chem. 1976, 122, 83. (b) Reyers, C.; Prock, A.; Giering, W. P. Organometallics 2002, 21, 546 (and references therein). (c) Poyatos, M.; Maisse-Franc-ois, A.; Bellemin-Laponnaz, S.; Gade, L. H. Organometallics 2006, 25, 2634–2641. (d) Tao, B.; Fu, G. C. Angew. Chem., Int. Ed. 2002, 41, 3892. (e) Faller, J. W.; Fonataine, P. P. Organometallics 2006, 25, 5887–5893. (f) Duan, W.-L.; Shi, M.; Rong, G.-B. Chem. Commun. 2003, 2916–2917. (g) Ito, H.; Kato, T.; Sawamura, M. Chem. Asian J. 2007, 2, 1436–1446. (h) Nishibayashi, Y.; Segawa, K.; Ohe, K.; Uemura, S. Organometallics 1995, 14, 5486–5487. (i) Chianese, A. R.; Crantree, R. H. Organometallics 2005, 24, 4432–4436. (16) (a) Gorsich, R. D. J. Am. Chem. Soc. 1960, 82, 4211. (b) Nifant'ev, I. E.; Butakov, K. A.; Aliev, Z. G.; Urazovskii, I. F. Metalloorg. Khim. 1991, 4, 1265. [ Organomet. Chem. USSR 1991, 4, 662 (Engl. Transl. )]. (c) A somewhat related Cp/X exchange was observed for manganocene: Kohler, F. H.; Hebendanz, N.; Thewalt, U.; Kanellakopulos, B.; Klenze, R. Angew. Chem., Int. Ed. Engl. 1984, 23, 721.
4502
Organometallics, Vol. 28, No. 15, 2009
Figure 1. Molecular structure of compound 2. Selected bond distances (in A˚) and angles (in deg): Nb1-Rh1 2.6744(2), Nb1N1 1.8643(17), Nb1-P1 2.4136(5), Nb1-Cl1 2.4349(5), Nb1Cp 2.67(6), Rh1-N1 2.0817(16), Rh1-P1 2.2516(5), Rh1-Cp 2.06(10), N1-Nb1-Cl1 106.71(5), P1-Nb1-Cl1 100.998(18), N1-Nb1-P1 103.08(5), N1-Rh1-P1 101.91(5). Hydrogen atoms are omitted for clarity.
X-ray structure analysis proved unequivocally the formulation of 2 as the Cp/Cl exchange product Cp((Cl)Nb(μNBut)(μ-PPh2)RhCp, but not the target complex Cp2Nb(dNBut)(μ-PPh2)RhCl(C2H4)2. The molecular structure of 2 exhibits a three-leg piano-stool geometry for niobium and a two-leg piano-stool geometry for rhodium; both metal atoms are bridged by the imido and phosphido ligands. Our search in the Cambridge Structural Database17 did not reveal any bimetallic rhodium complex with a μ2-bridging imido ligand; thus complex 2 is the first representative of this class. However, its Rh-N bond of 2.082(2) A˚ is comparable with Rh-imido bonds in trimetallic complexes M2Rh(μ3-NR), in which the imido ligand bridges three metal centers (1.9802.102 A˚).17 The Rh-P bond of 2.2516(5) A˚ in 2 is at the shorter end of a very wide spectrum of rhodium-phosphido distances (2.202-2.840 A˚). Similarly, the Nb-N distance of 1.864(2) A˚ is the shortest in the class of imido-bridged Nb complexes (range 1.952-2.135 A˚), and the Nb-P bond of 2.4136(5) is the shortest in the class of phosphido-bridged Nb complexes (range 2.487-2.717 A˚). These structural features may be the consequence of a very short Nb-Rh distance (2.6744(2) A˚), which indicates the presence of a direct NbRh bond (compare with the sum of metal radii, 2.774 A˚) and thus accounts for the diamagnetism of this complex. Complex 2 is stable in the solid state but gradually decomposes in solution at room temperature to give unidentified products. It does not react with H3SiPh over the course of 29 h. Attempted Cl/H exchange in 2 under the action of L-Selectride (Li[BH(CH(CH3)C2H5)3]) resulted in no reaction. In an attempt to remove the chloride and to stop the Cp/Cl exchange leading to 2, solutions of [Rh(μ-Cl)(C2H4)2]2 in THF were pretreated with L-Selectride before the addition of 1. Although a reaction occurred, accompanied by the (17) Cambridge Structural Database: Allen, F. H. Acta Cryst. 2002, B58, 380–388.
Leelasubcharoen et al.
formation of a black precipitate, the benzene-soluble part contained only the starting phosphide 1 and a small amount of Cp2Nb(NBut)Cl. Similar pretreatment of [Rh(μ-Cl)(C2H4)2]2 with Me2PhSiH affords a new, fluxional Cp compound characterized by the 1H NMR (C6D6) signal at 6.21 ppm. This product however decomposes overnight to furnish Cp2Nb(NBut)Cl. No signals were observed in the 31P NMR, suggesting that the phosphorus-containing co-product is either fluxional or insoluble in benzene. When [Rh(μ-Cl)(C2H4)2]2 was pretreated with PPh3, the fluxional complex [Rh(μ-Cl)(C2H4)(PPh3)]2 characterized by the Rh-coupled 31P signal at 53.9 ppm (dd, J(Rh-P) = 184.6 Hz, J(Rh-P) = 6.0 Hz) was formed. Subsequent 1:1 reaction of this product with 1 results in a fast formation of a mixture containing a new ethylene compound, Cp2Nb(dNBut)(μ-PPh2)Rh(PPh3)(C2H4)Cl (3), the exchange product 2, and Cp2Nb(NBut)Cl in the ratio 10:3.5:3.0. Complex 3 is unstable. After 1 h at room temperature, the signals due to 2 and 3 diminish, leaving Cp2Nb(NBut)Cl as the main niobium product. Complex 3 is characterized by its 1H NMR Cp signal at 6.06 ppm integrated as 10 protons, a broad ethylene signal at 2.13 ppm integrated as four protons, and a singlet at 0.66 ppm (But group) integrated as nine protons. The position of the latter signal suggests that the imido group remains terminal. Preparation of Bimetallic Complexes: Reaction with [(μCl)(Rh(COD)]2. Reaction of 1 with 0.5 equiv of [(μ-Cl)(Rh(COD)]2 (COD = 1,5 cyclooctadiene) in toluene afforded the target bimetallic complex Cp2Nb(dNBut)(μ-PPh2)Rh(Cl)(COD) (4) (Scheme 1). Complex 4 was isolated from toluene/hexane solution in the form of orange crystals of its hexane solvate (4 3 0.5hexane), and its structure was established by multinuclear NMR and X-ray diffraction analysis. The 1H NMR spectrum of 4 in C6D6 shows, in addition to other signals, a doublet at 6.14 ppm (J(P-H) = 1.2 Hz) for the Cp ring integrated as 10 protons, two broad singlets at 5.62 and 3.15 ppm for nonequivalent olefin protons integrated as four each, and a singlet at 0.66 ppm for the terminal But-imido group. The downfield-shifted olefin signal at 5.62 ppm is coupled in the 1H-31P HSQC spectrum with the broad phosphorus resonance at 26.3 ppm (d, J(PRh) = 114.2 Hz), which suggests the mutual trans arrangements of these groups. The upfield olefin signal at 3.15 ppm, therefore, comes from the other double bond of COD that is trans to the chloride ligand, which is a weaker trans ligand than phosphide. These spectroscopic features are consistent with the presence of a four-coordinate Rh center bound to niobium via the phosphide bridge. Complex 4 is stable in the solid state but gradually decomposes in solutions at room temperature (half-life approximately 4 days) to give Cp2Nb(NBut)Cl and the known complex [Rh2(μ-Cl)(μ-PPh2)(COD)2], identified by the 31P NMR.18 After a few days, the latter product partially rearranges to [Rh(μ-PPh2)(COD)]2.19 The X-ray structure analysis of 4 confirms the phosphidobridged bimetallic structure (Figure 2). The niobium atom is in a typical disubstututed niobocene geometry.20 The imido ligand remains terminal, forming a short Nb-N bond of 1.776(4) A˚, which is close to the mean NbdN bond of 1.782 A˚ found for the wide range of niobium-imido bonds (18) Burkhardt, E. W.; Mercer, W. C. Inorg. Chem. 1984, 23, 1779. (19) Kreter, P. E.; Meek, D. W. Inorg. Chem. 1983, 22, 319. (20) Lauher, J. W.; Hoffmann, R. J. Am. Chem. Soc. 1976, 98, 1729.
Article
Organometallics, Vol. 28, No. 15, 2009
4503
Scheme 1. Preparation of Complexes 4 and 5
Figure 2. Molecular structure of the compound 4 3 0.5hexane. Selected bond distances (in A˚) and angles (in deg): Nb1-P1 2.6854(13), Rh1-P1 2.3790(12), Nb1-N1 1.776(4), Rh1-Cl1 2.3968(13), Rh1-C24 2.102(5), Rh1-C23 2.137(5), Rh1-C28 2.183(5), Rh1-C27 2.209(4), N1-Nb1-P1 94.10(13), C31N1-Nb1 166.3(4), P1-Rh1-Cl1 92.27(4), C24-Rh1-P1 94.92(14), C23-Rh1-P1 93.28(13), C28-Rh1-P1 152.75(13), C27-Rh1-P1 170.35(14), C24-Rh1-Cl1 149.93(15), C23Rh1-Cl1 168.88(14), C28-Rh1-Cl1 89.02(14), C27-Rh1Cl1 86.80(14). Hydrogen atoms and the hexane solvate are omitted for clarity.
(1.672-2.051 A˚).17 For related niobocene imido complexes, a narrow range of NbdN distances has been determined (1.737(6)-1.804(5) A˚).12,21 The bond angle Nb-N-But of 166.2(3)° is also quite normal for niobocene imido complexes.21a The Nb-P bond of 2.685(1) A˚ and the Rh-P bond of 2.379(1) A˚ are noticeably longer than the corresponding distances in compound 2 discussed above, which can be accounted for by the absence of direct Nb-Rh bonding at the long Nb-Rh separation of 4.266 A˚. The Rh atom is ligated by the chloride ligand, two double bonds of COD, and the bridging phosphide, which together comprise the usual, square-planar geometry around the d8 Rh center. Consistent with the NMR data, one double bond of the (21) (a) Garces, A.; Perez, Y.; G omez-Ruiz, S.; Fajardo, M.; Anti~ nolo, A.; Otero, A.; L opez-Mardomingo, C.; G omez-Sal, P.; Prashar, S. J. Organomet. Chem. 2006, 691, 3652. (b) Anti~nolo, A.; Fajardo, M.; L opez-Mardomingo, C.; Lopez-Solera, I.; Otero, A.; Perez, Y.; Prashar, S. Organometallics 2001, 20, 3132. (c) Green, M. L. H.; James, J. T.; Chernega, A. N. J. Chem. Soc., Dalton Trans. 1997, 1719. (d) Herrman, W. A.; Baratta, W.; Herdtweck, E. Angew. Chem., Int. Ed. Engl. 1996, 35, 1951. (e) Chernega, A. N.; Green, M. L. H.; Suarez, A. G. Can. J. Chem. 1995, 73, 1157. (f) Schmidt, S.; Sundermeyer, J. J. Organomet. Chem. 1994, 472, 127.
COD ligand is positioned trans to the phosphide, and the other one is trans to the chloride ligand. The Rh-CCOD distances trans to the phosphide (2.183(5) and 2.209(4) A˚) are longer than those trans to the chloride (2.102(5) and 2.137(5) A˚), which highlights the stronger trans influence of the phosphide ligand. Catalytic Studies. On the basis of literature precedents on catalytic hydrosilylation of carbonyls by cationic Rh complexes,15b,15c,15e we initially sought to study the catalytic activity of the cationic complex [Cp2Nb(μ-NBut)(μ-PPh2)Rh(COD)]þ (5), which could be prepared by chloride abstraction from 4. We anticipated that the imido ligand would be able to occupy the bridging position, so that the fragment Cp2Nb(NBut)(PPh2) would serve as a hemilabile ligand to the cation [Rh(COD)]þ. The cation 5 was generated in CD2Cl2 by reacting 4 with AgBF4 (Scheme 1). Upon mixing the reagents, the color of the solution changed immediately from yellow to brown, accompanied by the formation of a brown precipitate. The 1 H NMR showed the formation of a new species, which however turned out to be too unstable to allow isolation. In light of this, hydrosilylation of acetophenone by PhSiH3 catalyzed by the in situ formed 5 was attempted. The course of the reaction in CD2Cl2 was monitored by 1H NMR spectroscopy. In 5 h, complete conversion was achieved, giving a mixture of mono- and bis(hydrosilylation) products (PhSiH2OCH(Me)Ph and PhSiH(OCH(Me)Ph)2, respectively) in the ratio 38:62 (Table 1). However, to our surprise, the parent neutral complex 4 showed even greater activity. The reaction was complete in 2 h, showing the same product composition (Table 1, entries 2 and 3). The change of solvent from methylene chloride to benzene had no effect on the rate of hydrosilylation or product distribution (Table 1, entries 2 and 3), whereas the nature of the silane does affect the reaction. With a secondary silane (PhMeSiH2), the reaction took more time and up to 30% of a dehydrogenative silylation product, PhSiHMeOC(Ph)dCH2, was formed (Table 1, entry 4). Tertiary silanes were not active in this reaction (Table 1, entries 5-7). For comparison, hydrosilylation of acetophenone was attempted with the parent Rh complex [Rh(μ-Cl)(COD)]2 in the absence of any added phosphide. [Rh(μ-Cl)(COD)]2 turned out to be a better catalyst than 4 for the hydrosilylation with PhMeSiH2 (Table 1, entry 4 vs entry 15), but a worse one for the hydrosilylation with PhSiH3 (Table 1, entry 14 vs entry 2). Even more surprisingly, the Cp/Cl exchange complex 2 also showed significant, although much reduced catalytic activity in the hydrosilylation of acetophenone (Table 1, entries 9-12). Again, the primary silane PhSiH3 performed much better than the secondary one (Table 1, entry 8 vs entry 9), while tertiary silanes were not active (entries 10-12). Complex 2 does not catalyze the addition of Me2PhSiH to 1-hexene. The isomerization to 2-hexene takes place instead.
4504
Organometallics, Vol. 28, No. 15, 2009
Leelasubcharoen et al.
Table 1. Catalytic Hydrosilylation of Acetophenone by 2, 4, and 5 (3%) in C6D6 entry
catalyst
silane
time
yield (%)
1
5
PhSiH3
5 ha
PhSiH2OCH(Me)Ph: 32% PhSiH(OCH(Me)Ph)2: 68%
2
4
PhSiH3
2 ha
3
PhSiH3
2h
4
MePhSiH2
5h
5 6 7
Me2PhSiH Et3SiH (OEt)3SiH
1d 1d 1d
PhSiH2OCH(Me)Ph: 40% PhSiH(OCH(Me)Ph)2: 60% PhSiH2OCH(Me)Ph: 32% PhSiH(OCH(Me)Ph)2: 68% PhSiHMeOCH(Me)Ph: 60% PhSiHMeOC(Ph)dCH2: 30% NR NR NR
PhSiH3
7 h 10 min
MePhSiH2 Me2PhSiH Et3SiH (OEt)3SiH
1 d 22 h 1d 1d 1d
8
2
9 10 11 12 14
[Rh(μ-Cl)(COD)]2
PhSiH3
7 h 30 min
15
[Rh(μ-Cl)(COD)]2
MePhSiH2
2h
a
PhSiH2OCH(Me)Ph: 20% PhSiH(OCH(Me)Ph)2: 80%. PhSiHMeOCH(Me)Ph 100% NR NR NR PhSiH2OCH(Me)Ph: 19% PhSiH(OCH(Me)Ph)2: 81% PhSiHMeOCH(Me)Ph: 63% PhSiHMeOC(Ph)dCH2: 34%.
In CD2Cl2.
Complex 4 was also found to catalyze hydrosilylation of benzaldehyde with PhSiH3 at a low catalyst load (1.5 mol %, 50 min for completeness), with the products being mono- and bis(addition) derivatives, PhSiH2OCH2Ph and PhSiH(OCH2Ph)2,22 in the ratio 36:64. At 3.0% catalyst load, the reaction was complete in 5 min, showing the selective formation of PhSiH(OCH2Ph)2 (94% yield). In an attempt to gain more insight into these catalytic transformations, stoichiometric reactions were attempted. When complex 4 reacts with PhSiH3 in a 1:1 ratio, decomposition occurs, leading to the formation of Cp2Nb(NBut)Cl and Cp2Nb(NBut)(PPh2). With a stoichiometric amount of PhCHO, 4 decomposes to Cp2Nb(NBut)Cl, but no new detectable Rh species could be identified in the reaction mixture by 1H and 31P NMR.
Conclusions The new imido phosphido complex Cp2Nb(NBut)(PPh2) (1) serves as a metallophosphine ligand to rhodium olefin fragments {ClRhL2}, affording phosphido-bridged bimetallic complexes. The outcome of the reaction depends on the nature of the olefin L. With L = ethylene, the main product is an unusual Cp/Cl exchange complex, Cp(Cl)Nb(μ-NBut)(μ-PPh2)RhCp (2), featuring a bridging imido group in addition to bridging phosphido group. When L2 is the chelating diolefin COD, the reaction of 1 with 0.5 equiv of [(μ-Cl)RhCOD]2 proceeds as a simple phosphide addition to furnish the bimetallic product Cp2Nb(dNBut)(μ-PPh2)Rh(Cl)(COD) (4). Both 2 and 4 were found to be active precatalysts for hydrosilylation of acetophenone and benzaldehyde. The neutral complex 4 exhibits enhanced reactivity in comparison with the cationic derivative [Cp2Nb(NBut)(PPh2)Rh(COD)]þ, presumably because the latter has the imido ligand too tightly coordinating to the Rh center. Future research will be directed at the optimization (22) (a) Yun, S. S.; Lee, J.; Lee, S. Bull. Korean Chem. Soc. 2001, 22, 623. (b) Yun, S. S.; Kin, T. S.; Kin, C. H. Bull. Korean Chem. Soc. 1994, 15, 522.
of the starting imido-phopshido complex of Nb to make it a better hemilabile metalloligand.
Experimental Details All manipulations were carried out, using conventional inert atmosphere glovebox and Schlenk techniques. Solvents were dried by Grubbs-type columns or by distillation from appropriate drying agents. NMR spectra were obtained with Bruker DPX-300 and Bruker DPX-600 instruments (1H: 300 and 600 MHz; 13C: 75.5 and 151 MHz; 29Si: 119.2 MHz; 31P: 121.5 and 243 MHz). IR spectra were measured on an ATI Mattson FTIR spectrometer. L-Selectride was purchased from Alfa Aesar, and [Rh(μ-Cl)(COD)]2 and [Rh(μ-Cl)(C2H4)]2 were from Strem. Silanes PhSiH3, MePhSiH, and MePh2SiH were prepared by reducing the corresponding chlorides with LiAlH4, and Et3SiH and (OEt)3SiH were obtained from Aldrich. Complexes Cp2Nb(NtBu)Cl12 and Ph2PLi 3 Et2O23 were prepared according to the literature methods. Synthesis of Cp2Nb(NBut)(PPh2) (1). Ph2PLi 3 Et2O (0.177 g, 0.665 mmol) dissolved in Et2O (20 mL) was added under stirring to a solution of Cp2Nb(NBut)Cl (0.213 g, 0.646 mmol) in Et2O (20 mL) at -30 °C. The color changed first from yellow to brown and then gradually turned to brown-green in the course of 2 h. The solution was allowed to warm to room temperature and then was stirred overnight, affording a brown suspension. After filtration through a microglass filter paper, volatiles were removed from the yellow solution in vacuo to give a yellow solid. The latter was washed with hexane and recrystallized from Et2O at -20 °C, affording a yellow microcrystalline material in 77% yield (0.245 g, 0.511 mmol). 1H NMR (300 MHz, C6D6, ppm): δ 1.02 (s, 9H, But), 5.52 (d, J(P-H) = 0.6 Hz, 10H, Cp), 7.05 (t, J(H-H) = 7.2 Hz, 2H, p-Ph), 7.19 (t, J(H-H) = 7.2 Hz, 4H, m-Ph), 7.80 (dd, J(H-H) = 7.2 Hz, J(P-H) = 9.3 Hz, 4H, o-Ph). 13C NMR (75.5 MHz, C6D6, ppm): δ 31.0 (But), 108.0 (d, J(P-C) = 2.3 Hz, Cp), 125.7 (p-Ph), 127.8 (m-Ph), 135.0 (d, J(C-P) = 16.6 Hz, o-Ph), 148.9 (d, J(C-P) = 35.4 Hz, i-Ph). 1 H-31P HSQC NMR (C6D6, ppm): δ -0.2 (s, PPh2). Anal. Calcd for C26H29NNbP: C, 65.14; H, 6.10; N, 2.92. Found: C, 65.04; H, 6.16; N, 2.80. (23) Bartlett, R. A.; Olmstead, M. M.; Power, P. P. Inorg. Chem. 1986, 25, 1243.
Article Synthesis of CpNbCl(μ-NBut)(μ-PPh2)RhCp (2). A solution of complex 1 (0.101 g, 0.211 mmol) in toluene (5 mL) was added to a stirred solution of [Rh(μ-Cl)(C2H4)2]2 (0.035 g, 0.105 mmol) in toluene (5 mL). The resulting dark green solution was stirred for 15 min, after which time the volatiles were removed in vacuo. The green solid produced was washed with hexane and recrystallized from Et2O, giving a green microcrystalline material in 81% yield (0.105 g, 0.170 mmol). 1H NMR (300 MHz, C6D6, ppm): δ 1.72 (s, 9H, But), 4.88 (s, 5H, Cp), 5.42 (d, J(P-H) = 0.6 Hz), 5H, Cp), 6.99 (m, 4H, m-Ph), 7.05 (m, 2H, p-Ph), 7.28 (m, 2H, o-Ph), 8.06 (dd, J(H-H) = 8.3 Hz, J(P-H) = 13.0 Hz, 2H, o-Ph). 13C NMR (75.5 MHz, C6D6, ppm): δ 35.4 (But), 87.3 (Cp), 104.4 (Cp), 133.2, 134.1 (Ph). Anal. Calcd for C26H29ClNNbPRh: C, 50.55; H, 4.73; N, 2.27. Found: C, 51.04; H, 5.25; N, 1.83. Generation of Cp2Nb(dNBut)(μ-PPh2)Rh(PPh3)(C2H4)Cl (3). To [Rh(μ-Cl)(C2H4)2]2 (0.003 g, 0.009 mmol) in C6D6 (0.6 mL) was added PPh3 (0.005 g, 0.019 mmol) to give an orange solution of [Rh(μ-Cl)(C2H4)(PPh3)]2 (31P NMR (121.5 MHz, C6D6, ppm): δ 53.9 ppm (dd, J(Rh-P) = 184.6 Hz, J(Rh-P) = 6.0 Hz)). Then 0.009 g (0.019 mmol) of complex Cp2Nb(NtBu)(PPh2) (1) was added to the mixture, causing the color change to brown. NMR spectra revealed the formation of a mixture of compounds containing Cp2Nb(dNBut)(μ-PPh2)Rh(PPh3)(C2H4)Cl (3), 2, and Cp2Nb(NBut)Cl in the ratio 10:3.5:3.0. Complex 3 decomposes in benzene solutions in the course of 1 h. Selected NMR data for Cp2Nb(dNBut)(μ-PPh2)Rh(PPh3)(C2H4)Cl (3): 1H NMR (300 MHz, C6D6, ppm): δ 6.06 (s, 10, H, Cp), 2.13 (bd, J(Rh-H) = 1.8 Hz, 4 H, C2H4), 0.66 (s, 9H, But). 31 P NMR (121.5 MHz, C6D6, ppm): δ 36.4 (d, J(Rh-P) = 128.8 Hz, PPh3). Synthesis of [Cp2NbCl(NBut)(μ-PPh2)RhCl(COD)] 3 0.5hexane (4 3 0.5hexane). A solution of complex 1 (0.100 g, 0.209 mmol) in toluene (5 mL) was added to a stirred solution of [Rh(μ-Cl)(COD)]2 (0.052 g, 0.106 mmol) in toluene (5 mL). The resulting orange solution was stirred for 15 min, after which time the solvent was removed in vacuo. The orange solid produced was washed with hexane and recrystallized by layer diffusion of hexane into a toluene solution, giving orange crystals of 4 3 0.5hexane in 91% yield (0.146 g, 0.190 mmol). 1H NMR (600 MHz, toluene-d8, 237 K, ppm): δ 0.69 (s, 9H, But), 0.95 (t, J(H-H) = 7.2 Hz, 3H, 0.5 hexane), 1.25 (m, 4H, 0.5 hexane) 1.65 (m, 2H, COD), 1.80 (m, 2H, COD), 2.34 (m, 4H, COD), 3.17 (s, 2H, COD), 5.62 (s, 4H, COD), 6.14 (s, 10H, Cp), 6.98 (m, 2H, p-Ph), 7.01 (m, 2H, Ph), 7.22 (m, 2H, Ph), 7.25 (m, 2H, Ph), 8.95 (bs, 2H, o-Ph). 1H NMR (300 MHz, 295 K, C6D6, ppm): δ 0.66 (s, 9H, But), 1.59 (m, 2, COD), 1.72, (m, 2H, COD), 2.27 (m, 4H, COD), 3.15 (bs, 2H, COD), 5.62 (bs, 2H, COD), 6.14 (d, J(P-H) = 1.2 Hz, 10H, Cp), 6.98 (t, J(H-H) = 7.4 Hz, 2H, p-Ph), 7.01 (t, J(H-H) = 7.4 Hz, 4H, m-Ph), 8.11 (bs, 4H, o-Ph). 1H NMR (300 MHz, 295 K, CD2Cl2, ppm): δ 0.81 (s, 9H,But), 1.77 (m, 2, COD), 1.92, (m, 2H, COD), 2.37 (m, 4H, COD), 2.93 (bs, 2H, COD), 5.09 (bs, 2H, COD), 6.21 (d, J(P-H) = 1.5 Hz, 10H, Cp), 7.17 (m, 2H, p-Ph), 7.25 (m, 4H, m-Ph), 7.8 (dd, J(H-H) = 7.5 Hz, J(H-P) = 7.5 Hz, 4H, o-Ph). 13C NMR (150.9 MHz, C6D6, 239 K, ppm): δ 14.5 (hexane), 23.2 (hexane), 29.4 (bs, COD), 30.7 (But), 32.2 (hexane), 33.0 (bs, COD), 70.7 (bs, COD), 98.3 (bs, COD), 110.0 (Cp), 126.8 (p-Ph), 127.4 (m, Ph), 127.7 (m-Ph), 129.2 (Ph), 135.1 (o-Ph), 143.7 (d, J = 12.9 Hz, i-Ph). 31P NMR (243.0 MHz, toluene-d8, 239 K, ppm): δ 26.3 (d, J(Rh-P) = 114.2 Hz). Anal. Calcd for C37H48ClNNbPRh: C, 57.79; H, 6.29; N, 1.82. Found: C, 57.85: H, 6.18; N, 1.83. General Procedure for the Hydrosilylation of PhC(O)CH3 and PhC(O)H Catalyzed by 2 or 4 in Benzene. The catalyst (0.003 mmol, 2.3 mol %) was dissolved in 0.6 mL of C6D6. (24) (a) Yun, S. S.; Lee, J.; Lee, S. Bull. Korean Chem. Soc. 2001, 22, 623. (b) Yun, S. S.; Kim, T. S.; Kim, C. H. Bull. Korean Chem. Soc. 1994, 15, 522. (c) Peterson, E.; Khalimon, A. Y.; Simionescu, R.; Kuzmina, L. G.; Howard, J. A. K.; Nikonov, G. I. J. Am. Chem. Soc. 2009, 131, 908.
Organometallics, Vol. 28, No. 15, 2009
4505
Table 2. X-ray Crystal Data and Structure Refinement for 2 and 4 3 0.5hexane formula fw color, habit cryst size, mm cryst syst space group a, A˚ b, A˚ c, A˚ β, deg V, A˚3 Z T, °C Fcalc, g/cm3 F(000) μ, mm-1 2θmax, deg total no. of reflns unique reflns Rint no. with I g 2σ(I) no. of variables R1 (I g 2σ(I)) wR2 (all data) GOOF
2
4 3 0.5hexane
C26H29ClNNbPRh 617.74 green, needle 0.20 0.08 0.08 monoclinic P21/n 9.1504(3) 16.5399(6) 16.5001(6) 103.540(1) 2427.83(15) 4 150 1.690 1240 1.342 58.00 26 584 6445 0.0324 5497 280 0.0239 0.0586 1.050
C37H48ClNNbPRh 769.00 yellow-orange, block 0.38 0.12 0.10 monoclinic P21/c 9.1766(6) 18.6877(12) 20.1635(12) 98.312(3) 3421.5(4) 4 123 1.493 1580 0.968 56.00 23 855 8234 0.0531 6317 379 0.0587 0.1113 1.164
The silane (16 μL, 0.130 mmol) was added into the solution, followed by the carbonyl compound (0.128 mmol). The course of the reaction was monitored by 1H NMR spectroscopy. The hydrosilylation products PhSiH2OCH2Ph,24 PhSiH(OCH2Ph)2,24 PhSiH2(OCH(Me)Ph),25 and PhSiH(OCH(Me)Ph)225 were characterized by NMR. First Stereoisomer of PhSiHMeOCH(Me)Ph. 26 1H NMR (300 MHz, C6D6, ppm): δ 0.30 (d, 3H, SiMe), 1.38 (d, 3H, Me), 5.29 (q, 1H, SiH), 7.61 (m, SiPh). 1H-29Si HSQC NMR (1H: 300 MHz; 29Si: 59.6 MHz, C6D6, J = 7 Hz, ppm): δ -4. 29Si INEPTþ NMR (119.2 MHz, C6D6, J = 200 Hz, ppm): δ -4.3 (d, 1JSi-H = 210). 1H-13C HSQC NMR (1H: 300 MHz; 13C: 75.5 MHz, C6D6, J = 145 Hz, ppm): δ -1.0 (OMePh), 22.0 (SiMe). Second Stereoisomer of PhSiHMeOCH(Me)Ph. 1H NMR (300 MHz, C6D6, ppm): δ 0.33 (d, 3H, Me), 1.41 (d, 3H, MeSi), 5.42 (q, 1H, SiH), 7.52 (m, SiPh). 1H-29Si HSQC NMR (1H: 300 MHz; 29Si: 59.6 MHz, C6D6, J = 7 Hz, ppm): δ -4.8 (d, 1 JSi-H = 208 Hz). 1H-13C HSQC NMR (1H: 300 MHz; 13C: 75.5 MHz, C6D6, J = 145 Hz, ppm): δ -1.0 (OMePh), 22.0 (SiMe). PhSiHMeOC(Ph)dCH2. 27 1H NMR (300 MHz, C6D6, ppm): δ 0.40 (d, 3H, SiMe), 4.57 (d, 1H, CdCH2), 4.90 (d, 1H, CdCH2), 5.47 (q, 1H, PhSiHMe). 1H-29Si HSQC NMR (1H: 300 MHz; 29Si: 59.6 MHz, C6D6, J = 7 Hz, ppm): δ -22. 1 H-13C HSQC NMR (1H: 300 MHz; 13C: 75.5 MHz, C6D6, J = 145 Hz, ppm): δ 90 (CdCH2). General Procedure for the Hydrosilylation of PhC(O)CH3 by PhSiH3 Catalyzed by 4 or 4/AgBF4 in CD2Cl2. Complex 4 (4.0 mg, 0.005 mmol, 3.0 mol %) or a mixture of 4 (4.0 mg, 0.005 mmol, 3.0 mol %) and AgBF4 (1.0 mg, 0.005 mmol) was dissolved in 0.6 mL of CD2Cl2. PhSiH3 (23 μL, 0.186 mmol) and PhC(O)CH3 (22.0 μL, 0.188 mmol) were charged into the mixture. The course of the reaction was monitored by 1H NMR spectroscopy. (25) (a) Khalimon, A. Y.; Simionescu, R.; Kuzmina, L. G.; Howard, J. A. K.; Nikonov, G. I. Angew. Chem., Int. Ed. 2008, 47, 7701. (b) Yun, S. S.; Yang, Y. S.; Lee, S. Bull. Korean Chem. Soc. 1997, 18, 1058. (26) (a) Glushkova, N. E.; Kochina, T. A.; Kharitonov, N. P. Zhur. Obshch. Khim. 1984, 54, 2741. (b) Tamio, H.; Yamamoto, K.; Kumada, M. J. Organomet. Chem. 1976, 112, 253. (27) Iretskii, A. V.; Zhidkova, O. B.; Dolgushina, T. S.; Labeish, N. N.; Skvortsova, N. K.; Galishev, V. A. Zhur. Obshch. Khim. 1993, 63, 1681.
4506
Organometallics, Vol. 28, No. 15, 2009
Leelasubcharoen et al.
Crystal Structure Determinations. X-ray quality crystals of 2 were obtained by low-temperature crystallization from ether, and crystals of 4 3 0.5hexane were grown by layer diffusion of hexane into a toluene solution of 4. In both cases, crystals were covered by polyperfluoro oil and mounted directly onto the Bruker Smart three-circle diffractometer with CCD area detector at 150 and 123 K, respectively. Experimental intensities were collected using Mo KR radiation (0.71073 A˚) in ω-scan mode. The crystallographic data and characteristics of structure solution and refinement are assembled in Table 2. The Bruker SAINT program28 was used for data reduction. The structures
were solved by direct methods29 and refined by full-matrix leastsquares procedures, using w(|F2o| - |F2c |)2 as the refined function. In both structures, all non-hydrogen atoms were refined with anisotropic thermal parameters and all hydrogen atoms were placed in calculated positions and refined using a riding model.
(28) SAINT, Version 6.02A; Bruker AXS Inc.: Madison, WI, 2001. (29) SHELXTL-Plus, Release 5.10; Bruker AXS Inc.: Madison, WI, 1997.
Supporting Information Available: Crystallographic information files (CIF) for compounds 2 and 4. This material is available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. This research was supported by Research Corporation and NSERC grants to G.I.N. L. G.K. and J.A.K.H. thank the Royal Society of Chemistry and the University of Durham for support.
